commit 0c2e763e2a205e51f63f4e797ca6e2f593f4693f
parent 0767ed15fbcebd38536803c958cf6ef0e3334046
Author: Christian Grothoff <christian@grothoff.org>
Date: Sat, 16 Sep 2006 18:20:48 +0000
fx
Diffstat:
9 files changed, 109 insertions(+), 59 deletions(-)
diff --git a/ChangeLog b/ChangeLog
@@ -1,3 +1,7 @@
+Sat Sep 16 12:36:42 MDT 2006
+ Added support for various additional tags to ID3v2 extractors.
+ Now (again) trimming whitespace at the end of ID3v1 tags.
+
Wed Sep 6 13:38:55 PDT 2006
Added tIME support to PNG extractor. Bugfixes in PDF extractors.
Made libextractor relocateable (plugin path no longer hardwired
diff --git a/configure.ac b/configure.ac
@@ -1,8 +1,8 @@
# Process this file with autoconf to produce a configure script.
AC_PREREQ(2.57)
-AC_INIT([libextractor], [0.5.15], [bug-libextractor@gnu.org])
+AC_INIT([libextractor], [0.5.15a], [bug-libextractor@gnu.org])
AC_REVISION($Revision: 1.67 $)
-AM_INIT_AUTOMAKE([libextractor], [0.5.15])
+AM_INIT_AUTOMAKE([libextractor], [0.5.15a])
AM_CONFIG_HEADER(src/include/config.h)
AH_TOP([#define _GNU_SOURCE 1])
diff --git a/src/include/extractor.h b/src/include/extractor.h
@@ -168,6 +168,19 @@ typedef enum {
EXTRACTOR_INFORMATION = 112,
EXTRACTOR_FULL_NAME = 113,
EXTRACTOR_CHAPTER = 114,
+ EXTRACTOR_YEAR = 115,
+
+ EXTRACTOR_LINK = 116,
+ EXTRACTOR_MUSIC_CD_IDENTIFIER = 117,
+ EXTRACTOR_PLAY_COUNTER = 118,
+ EXTRACTOR_POPULARITY_METER = 119,
+ EXTRACTOR_CONTENT_TYPE = 120,
+ EXTRACTOR_ENCODED_BY = 121,
+ EXTRACTOR_TIME = 122,
+ EXTRACTOR_MUSICIAN_CREDITS_LIST = 123,
+ EXTRACTOR_MOOD = 124,
+
+
} EXTRACTOR_KeywordType;
/**
diff --git a/src/main/extract.c b/src/main/extract.c
@@ -1,6 +1,6 @@
/*
This file is part of libextractor.
- (C) 2002, 2003, 2004, 2005 Vidyut Samanta and Christian Grothoff
+ (C) 2002, 2003, 2004, 2005, 2006 Vidyut Samanta and Christian Grothoff
libextractor is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published
@@ -594,39 +594,34 @@ main (int argc, char *argv[])
extractors = EXTRACTOR_addLibraryLast(extractors,
"libextractor_split");
- if (verbose == YES)
- {
- /* print list of all used extractors */
- }
/* extract keywords */
if ( bibtex == YES )
fprintf(stdout,
_("%% BiBTeX file\n"));
- for (i = optind; i < argc; i++)
- {
- errno = 0;
- keywords = EXTRACTOR_getKeywords (extractors, argv[i]);
- if (0 != errno) {
- if (verbose == YES) {
- fprintf(stderr,
- "%s: %s: %s\n",
- argv[0], argv[i], strerror(errno));
- }
- ret = 1;
- continue;
+ for (i = optind; i < argc; i++) {
+ errno = 0;
+ keywords = EXTRACTOR_getKeywords (extractors, argv[i]);
+ if (0 != errno) {
+ if (verbose == YES) {
+ fprintf(stderr,
+ "%s: %s: %s\n",
+ argv[0], argv[i], strerror(errno));
}
- if (duplicates != -1 || bibtex == YES)
- keywords = EXTRACTOR_removeDuplicateKeywords (keywords, duplicates);
- if (verbose == YES && bibtex == NO)
- printf (_("Keywords for file %s:\n"), argv[i]);
- if (bibtex == YES)
- printSelectedKeywordsBibtex (stdout, keywords, print, argv[i]);
- else
- printSelectedKeywords (stdout, keywords, print, verbose);
- if (verbose == YES && bibtex == NO)
- printf ("\n");
- EXTRACTOR_freeKeywords (keywords);
+ ret = 1;
+ continue;
}
+ if ( (duplicates != -1) || (bibtex == YES))
+ keywords = EXTRACTOR_removeDuplicateKeywords (keywords, duplicates);
+ if (verbose == YES && bibtex == NO)
+ printf (_("Keywords for file %s:\n"), argv[i]);
+ if (bibtex == YES)
+ printSelectedKeywordsBibtex (stdout, keywords, print, argv[i]);
+ else
+ printSelectedKeywords (stdout, keywords, print, verbose);
+ if (verbose == YES && bibtex == NO)
+ printf ("\n");
+ EXTRACTOR_freeKeywords (keywords);
+ }
free (print);
EXTRACTOR_removeAll (extractors);
diff --git a/src/main/extractor.c b/src/main/extractor.c
@@ -1,6 +1,6 @@
/*
This file is part of libextractor.
- (C) 2002, 2003, 2004, 2005 Vidyut Samanta and Christian Grothoff
+ (C) 2002, 2003, 2004, 2005, 2006 Vidyut Samanta and Christian Grothoff
libextractor is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published
@@ -151,16 +151,26 @@ static const char *keywordTypes[] = {
gettext_noop("editing cycles"),
gettext_noop("scale"),
gettext_noop("manager"),
- gettext_noop(/* movie director */"director"),
+ gettext_noop(/* movie director */"director"), /* 110 */
gettext_noop("duration"),
gettext_noop("information"),
gettext_noop("full name"),
gettext_noop("chapter"),
+ gettext_noop("year"), /* 115 */
+ gettext_noop("link"),
+ gettext_noop("music CD identifier"),
+ gettext_noop("play counter"),
+ gettext_noop("popularity meter"),
+ gettext_noop("content type"), /* 120 */
+ gettext_noop("encoded by"),
+ gettext_noop("time"),
+ gettext_noop("musician credits list"),
+ gettext_noop("mood"),
NULL,
};
/* the number of keyword types (for bounds-checking) */
-#define HIGHEST_TYPE_NUMBER 115
+#define HIGHEST_TYPE_NUMBER 125
#ifdef HAVE_LIBOGG
#if HAVE_VORBIS
@@ -1236,11 +1246,11 @@ removeKeyword (const char *keyword,
}
if (pos == NULL)
break;
- if ((0 == strcmp (pos->keyword, keyword)) &&
- ((pos->keywordType == type) ||
- (((options & EXTRACTOR_DUPLICATES_TYPELESS) > 0)) ||
- (((options & EXTRACTOR_DUPLICATES_REMOVE_UNKNOWN) > 0) &&
- (pos->keywordType == EXTRACTOR_UNKNOWN))))
+ if ( (0 == strcmp (pos->keyword, keyword)) &&
+ ( (pos->keywordType == type) ||
+ (((options & EXTRACTOR_DUPLICATES_TYPELESS) > 0)) ||
+ ( ((options & EXTRACTOR_DUPLICATES_REMOVE_UNKNOWN) > 0) &&
+ (pos->keywordType == EXTRACTOR_UNKNOWN)) ) )
{
/* remove! */
if (prev == NULL)
@@ -1269,16 +1279,18 @@ removeKeyword (const char *keyword,
*/
EXTRACTOR_KeywordList *
EXTRACTOR_removeDuplicateKeywords (EXTRACTOR_KeywordList * list,
- const unsigned int options)
-{
+ const unsigned int options) {
EXTRACTOR_KeywordList *pos;
pos = list;
- while (pos != NULL)
- {
- removeKeyword (pos->keyword, pos->keywordType, options, &list, pos);
- pos = pos->next;
- }
+ while (pos != NULL) {
+ removeKeyword(pos->keyword,
+ pos->keywordType,
+ options,
+ &list,
+ pos);
+ pos = pos->next;
+ }
return list;
}
@@ -1288,8 +1300,7 @@ EXTRACTOR_removeDuplicateKeywords (EXTRACTOR_KeywordList * list,
* @return a list of keywords without duplicates
*/
EXTRACTOR_KeywordList *
-EXTRACTOR_removeEmptyKeywords (EXTRACTOR_KeywordList * list)
-{
+EXTRACTOR_removeEmptyKeywords (EXTRACTOR_KeywordList * list) {
EXTRACTOR_KeywordList * pos;
EXTRACTOR_KeywordList * last;
diff --git a/src/plugins/id3v23extractor.c b/src/plugins/id3v23extractor.c
@@ -1,6 +1,6 @@
/*
This file is part of libextractor.
- (C) 2002, 2003, 2004 Vidyut Samanta and Christian Grothoff
+ (C) 2002, 2003, 2004, 2006 Vidyut Samanta and Christian Grothoff
libextractor is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published
@@ -55,11 +55,16 @@ typedef struct {
static Matches tmap[] = {
{ "COMM", EXTRACTOR_COMMENT },
+ { "IPLS", EXTRACTOR_CONTRIBUTOR },
+ { "LINK", EXTRACTOR_LINK },
+ { "MCDI", EXTRACTOR_MUSIC_CD_IDENTIFIER },
+ { "PCNT", EXTRACTOR_PLAY_COUNTER },
+ { "POPM", EXTRACTOR_POPULARITY_METER },
{ "TCOP", EXTRACTOR_COPYRIGHT },
{ "TDAT", EXTRACTOR_DATE },
- { "TCON", EXTRACTOR_GENRE },
+ { "TCON", EXTRACTOR_CONTENT_TYPE },
{ "TIT1", EXTRACTOR_GENRE },
- { "TENC", EXTRACTOR_PRODUCER },
+ { "TENC", EXTRACTOR_ENCODED_BY },
{ "TEXT", EXTRACTOR_LYRICS },
{ "TOLY", EXTRACTOR_CONTRIBUTOR },
{ "TOPE", EXTRACTOR_CONTRIBUTOR },
@@ -70,7 +75,7 @@ static Matches tmap[] = {
{ "TPE4", EXTRACTOR_INTERPRET },
{ "TMED", EXTRACTOR_MEDIA_TYPE },
{ "TCOM", EXTRACTOR_CREATOR },
- { "TIME", EXTRACTOR_DATE },
+ { "TIME", EXTRACTOR_TIME },
{ "TOFN", EXTRACTOR_FILENAME },
{ "TOPE", EXTRACTOR_ARTIST },
{ "TPUB", EXTRACTOR_PUBLISHER },
@@ -80,6 +85,8 @@ static Matches tmap[] = {
{ "TOAL", EXTRACTOR_ALBUM },
{ "TALB", EXTRACTOR_ALBUM },
{ "TLAN", EXTRACTOR_LANGUAGE },
+ { "TYER", EXTRACTOR_YEAR },
+ { "TLEN", EXTRACTOR_DURATION },
{ "TIT2", EXTRACTOR_TITLE },
{ "TIT3", EXTRACTOR_DESCRIPTION },
{ "WCOM", EXTRACTOR_RELEASE },
diff --git a/src/plugins/id3v24extractor.c b/src/plugins/id3v24extractor.c
@@ -1,6 +1,6 @@
/*
This file is part of libextractor.
- (C) 2002, 2003, 2004 Vidyut Samanta and Christian Grothoff
+ (C) 2002, 2003, 2004, 2006 Vidyut Samanta and Christian Grothoff
libextractor is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published
@@ -56,11 +56,19 @@ typedef struct {
static Matches tmap[] = {
{ "COMM", EXTRACTOR_COMMENT },
+ { "IPLS", EXTRACTOR_CONTRIBUTOR },
+ { "TIPL", EXTRACTOR_CONTRIBUTOR },
+ { "TMOO", EXTRACTOR_MOOD },
+ { "TMCL", EXTRACTOR_MUSICIAN_CREDITS_LIST },
+ { "LINK", EXTRACTOR_LINK },
+ { "MCDI", EXTRACTOR_MUSIC_CD_IDENTIFIER },
+ { "PCNT", EXTRACTOR_PLAY_COUNTER },
+ { "POPM", EXTRACTOR_POPULARITY_METER },
{ "TCOP", EXTRACTOR_COPYRIGHT },
{ "TDRC", EXTRACTOR_DATE },
{ "TCON", EXTRACTOR_GENRE },
{ "TIT1", EXTRACTOR_GENRE },
- { "TENC", EXTRACTOR_PRODUCER },
+ { "TENC", EXTRACTOR_ENCODED_BY },
{ "TEXT", EXTRACTOR_LYRICS },
{ "TOLY", EXTRACTOR_CONTRIBUTOR },
{ "TOPE", EXTRACTOR_CONTRIBUTOR },
@@ -69,6 +77,7 @@ static Matches tmap[] = {
{ "TPE2", EXTRACTOR_ARTIST },
{ "TPE3", EXTRACTOR_CONDUCTOR },
{ "TPE4", EXTRACTOR_INTERPRET },
+ { "TIME", EXTRACTOR_TIME },
{ "TMED", EXTRACTOR_MEDIA_TYPE },
{ "TCOM", EXTRACTOR_CREATOR },
{ "TOFN", EXTRACTOR_FILENAME },
@@ -77,6 +86,7 @@ static Matches tmap[] = {
{ "TRSN", EXTRACTOR_SOURCE },
{ "TRSO", EXTRACTOR_CREATED_FOR },
{ "TSRC", EXTRACTOR_RESOURCE_IDENTIFIER },
+ { "TYER", EXTRACTOR_YEAR },
{ "TOAL", EXTRACTOR_ALBUM },
{ "TALB", EXTRACTOR_ALBUM },
{ "TLAN", EXTRACTOR_LANGUAGE },
diff --git a/src/plugins/id3v2extractor.c b/src/plugins/id3v2extractor.c
@@ -1,6 +1,6 @@
/*
This file is part of libextractor.
- (C) 2002, 2003, 2004 Vidyut Samanta and Christian Grothoff
+ (C) 2002, 2003, 2004, 2006 Vidyut Samanta and Christian Grothoff
libextractor is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published
diff --git a/src/plugins/mp3extractor.c b/src/plugins/mp3extractor.c
@@ -1,6 +1,6 @@
/*
This file is part of libextractor.
- (C) 2002, 2003, 2004 Vidyut Samanta and Christian Grothoff
+ (C) 2002, 2003, 2004, 2006 Vidyut Samanta and Christian Grothoff
libextractor is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published
@@ -243,6 +243,12 @@ int freq_table[4][3]={
#define SYSERR 1
#define INVALID_ID3 2
+static void trim(char * k) {
+ while ( (strlen(k) > 0) &&
+ (isspace(k[strlen(k)-1])) )
+ k[strlen(k)-1] = '\0';
+}
+
static int get_id3(const char * data,
size_t size,
id3tag * id3) {
@@ -261,22 +267,27 @@ static int get_id3(const char * data,
id3->title = convertToUtf8(pos,
30,
"ISO-8859-1");
+ trim(id3->title);
pos += 30;
id3->artist = convertToUtf8(pos,
30,
"ISO-8859-1");
+ trim(id3->artist);
pos += 30;
id3->album = convertToUtf8(pos,
30,
"ISO-8859-1");
+ trim(id3->album);
pos += 30;
id3->year = convertToUtf8(pos,
4,
"ISO-8859-1");
+ trim(id3->year);
pos += 4;
id3->comment = convertToUtf8(pos,
30,
"ISO-8859-1");
+ trim(id3->comment);
pos += 30;
id3->genre = "";
if (pos[0] < GENRE_NAME_COUNT)
@@ -289,9 +300,8 @@ static struct EXTRACTOR_Keywords *
addkword(EXTRACTOR_KeywordList *oldhead,
const char * phrase,
EXTRACTOR_KeywordType type) {
-
EXTRACTOR_KeywordList * keyword;
-
+
keyword = malloc(sizeof(EXTRACTOR_KeywordList));
keyword->next = oldhead;
keyword->keyword = strdup(phrase);
@@ -460,10 +470,10 @@ libextractor_mp3_extract(const char * filename,
if (strlen(info.album) > 0)
klist = addkword(klist, info.album, EXTRACTOR_ALBUM);
if (strlen(info.year) > 0)
- klist = addkword(klist, info.year, EXTRACTOR_DATE);
+ klist = addkword(klist, info.year, EXTRACTOR_YEAR);
if (strlen(info.genre) > 0)
klist = addkword(klist, info.genre, EXTRACTOR_GENRE);
- if (strlen(info.genre) > 0)
+ if (strlen(info.comment) > 0)
klist = addkword(klist, info.comment, EXTRACTOR_COMMENT);
